Hydrodynamic characteristics in an inverse internal-loop airlift-driven fibrous-bed bioreactor

An inverse internal loop airlift-driven fibrous bed bioreactor (ALFBB) was designed by combining the advantages of an internal loop airlift bioreactor and packed bed bioreactor into one column. This bioreactor, with a high degree of design flexibility, is expected to handle genetically engineered cells as well as fragile cells, which are shear-sensitive. The hydrodynamic characteristics of the combined system have been investigated. Woven cotton was set in the downcomer of the I-IL-ALB to represent the fibrous bed packed bed and the outcome results were compared with those of the polyurethane foam (PUF) packed system and the unpacked I-IL-ALB system. The effects of the packing nature, packing height, packing top and bottom clearances, gaps between adjacent fiber surfaces, and superficial gas velocities were investigated. The hydrodynamic output variables included the gas holdup and liquid circulation velocity. Gas holdup for all packed systems continuously increased with increases in packing height, packing top clearance and superficial gas velocity. It was found highest in the downcomer of the cotton packed system than in the PUF counter part due to the roughness and hydrophilicity of the woven cotton fibrous material. Increased amounts of packing in the I-IL-ALB, whether in the form of cotton or PUF decreased the liquid circulation velocity in the bioreactor because of the increased frictional resistance and tortuosity. The reduction in liquid circulation velocity was significant for large packing with small gaps between fiber surfaces and increased bottom clearances of the cotton packed system. Empirical models based on packing properties are presented which accurately predict the gas holdup, whereas energy based model was proposed to predict liquid circulation velocities. The optimum hydrodynamic conditions were observed with cotton packing.     

Influence of top‐section design and draft‐tube height on the performance of airlift bioreactors containing water‐in‐oil microemulsion

The mixing and mass transfer characteristics of draft‐tube airlift bioreactors (DTAB) for a water‐in‐kerosene microemulsion, as a cold model of petroleum biodesulfurization, were studied. Incomplete gas disengagement at the top‐section of the DTAB and hence high gas recirculation were obtained with the microemulsion system for all the top‐section configurations employed in the present study especially at the high airflow rates. The ratio (S) of the volumes of the riser and the downcomer to the top‐section together with the gas disengagement abilities of the gas separator were both found to affect the mixing performance of the DTAB employed for the microemulsion system. Increase in the draft‐tube height resulted in significant increase in the mixing time (t_m) and a slight increase in the overall volumetric oxygen transfer coefficient (k_La). Increase in the diameter of the top‐section and the height of the liquid above the draft‐tube led to a decrease in k_La, the latter effect being less prominent. New correlations were developed that predicted the mixing time and oxygen transfer coefficients obtained in the present work with reasonable accuracy. Copyright © 2004 Society of Chemical Industry     

Does the fluid elasticity influence the dispersion in packed beds?

Reasons are given why the axial dispersion in a gas flowing through a packed bed may be influenced by the elasticity - or compressibility - of the fluid. To support this hypothesis, experiments have been done in a packed column at pressures from 0.13 to 2.0 MPa. The elasticity E of a gas is proportional to the pressure P and the compressibility to 1/P. The axial dispersion coefficients as determined were found to be a function of the pressure in the packed bed in the turbulent flow region of 3 < Re_p < 150 if the Bodenstein number is plotted as a function of the particle Reynolds number. This is shown to be an artifact. The pressure influence is eliminated, if Bo_{m, ax} is plotted versus the ratio of the kinetic forces over the elastic forces ?u²/E. Regrettably, Bo_{m, ax} seems to be independent of ?u²/E. For the moment we only can conclude that Bo_{m, ax} in the turbulent region is a unique function of the velocity of the gas which flows through the packed bed. Although the fact that a constant Bo value is obtained when plotted against ?u²/E, the experimental results are so intriguing we wanted to make them public already now. The experimental work proceeds.     

Gas‐liquid mass transfer in low‐ and medium‐consistency pulp suspensions

The volumetric gas—liquid mass transfer coefficient (k_La) was measured for low‐ and medium‐consistency pulp suspensions using the cobalt‐catalyzed sulfite oxidation technique. Mass transfer rates were measured in a high‐shear mixer for a range of operating parameters, including the rotor speed (N = 10 to 50 rev/s), gas void fraction (X_g = 0.10 to 0.40) and fibre mass concentration (C_m = 0.0 to 0.10). k_La measurements were compared with the macroscale flow regime in the vessel (characterized using photographic techniques) and correlated with energy dissipation, gas void fraction and suspension mass concentration in the mixer. We found that gas‐liquid mass transfer was significantly reduced in pulp suspensions, even for low suspension concentrations. Part of this reduction was associated with dissolved components leached from the fibres into the liquid phase. This could account for reductions in k_La of up to 30% when compared with distilled water. The fibres further reduced k_La, with the magnitude of the decrease depending on the fibre mass concentration. Correlations were developed for k_La and compared with results available in the literature.     

Automated measurements of gas‐liquid mass transfer in micropacked bed reactors

Gas‐liquid mass transfer in micropacked bed reactors is characterized with an automated platform integrated with in‐line Fourier transform infrared spectroscopy. This setup enables screening of a multidimensional parameter space underlying absorption with chemical reaction. Volumetric gas‐liquid mass‐transfer coefficients (k_La) are determined for the model reaction of CO₂ absorption in a methyl diethanolamine/water solution. Parametric studies are conducted varying gas and liquid superficial velocities, packed bed dimensions and packing particle sizes. The results show that k_La values are in the range of 0.12～0.39 s^?1, which is about one‐to‐two orders of magnitude larger than those of conventional trickle beds. An empirical correlation predicts k_La in micropacked bed reactors in good agreement with experimental data. © 2017 American Institute of Chemical Engineers AIChE J, 64: 564–570, 2018     

Determination of mass transfer coefficients for packing materials used in biofilters and biotrickling filters for air pollution control—2: Development of mass transfer coefficients correlations

Correlations that allow determination of gas film mass transfer coefficients (k_Ga_t, k_Ga_w) and liquid film mass transfer coefficients (k_La_w) for packing materials used in biofilters and biotrickling filters for air pollution control are presented. Lava rock, polyurethane foam cubes (PUF), Pall rings, porous ceramic beads, porous ceramic Raschig rings, and various compost-woodchips mixtures were used as packing materials. The functionality of gas and liquid velocity on mass transfer coefficients (k_Ga_t,k_Ga_w,k_La_w) obtained experimentally (see Part 1 of this paper) was correlated using modified Onda-type equations. The correlation equations helped to better understand the sensitivity of gas and liquid velocities on mass transfer, and the effects of packing wetting. Each packing had a different functionality with gas and liquid velocity and different wetting property, hence different correlation equations were needed for the different packing materials. Most of the fitted data fell within ±20% of the experimental values.     

Hydrodynamic and mass transfer characteristics of bubble and packed bubble columns with downcomer

The hydrodynamic and mass transfer characteristics of bubble and packed bubble columns with downcomer were investigated. The contactor consisted of two concentric columns of 0.11 and 0.2 m i.d., with the annulus acting as the downcomer. The packing used in this investigation was standard 16 mm stainless steel Pall rings. The superficial gas and liquid velocities, V_G and V_L, were varied from 0.01 to 0.09 and 1 × 10^?3 to 8.8 × 10^?3 m s^?1 respectively. Two flow patterns, namely the bubble and pulse flows were observed in the packed bubble column with downcomer, as shown by a flow map. The liquid circulation velocity in both the contactors was observed to be constant throughout the ranges of V_G and V_L covered in this work. The effect of liquid viscosity (0.8 to 9.5 mPa ? s) and surface tension (45 to 72 mN m^?1) on the flow pattern, liquid circulation, gas hold-up and pressure drop was investigated. The pressure drop characteristics across the two contactors have been compared with those across a bubble column. Values of the effective interfacial area, a, and the volumetric mass transfer coefficient, k_L a, were measured by using chemical methods. Values of a as high as 180 and 700 m^?1 and k_L a as high as 0.075 and 0.22 s^?1, in the bubble and packed bubble columns with downcomer, respectively, were obtained. The values of true liquid-side mass transfer coefficient, k_L, were found to be independent of V_G and were of the order of 5.5 × 10^?4 and 3.5 × 10^?4 m s^?1, respectively, in the two contactors.     

Gas mixing in fluidized beds containing screen packings

Gas mixing was studied in a fluidized bed containing hollow, open-wall packings using a continuous injection of oxygen as a tracer. During bed expansion, the reciprocal of mixing length, u/E₁, for a gas-fiuidized bed containing packings of dimension less than 3-in. was found to decrease and then increase with the expansion ratio, Δ, passing through a minimum, in contrast to an unpacked fluidized bed for which no minimum u/E₁, was reported. The minimum u/E₁, occurred always at Δ = 0.30, independent of column diameter, of the shape or of the size of the packing, and of the size of the fluidizing particle. Values of the minimum u/E₁'s for a fluidized bed packed with various screen cylinders depended upon the size of the packing, d_n, and of the fluidized particle, d₁,. An empirical correlation of (u/E₁)mn, with d₁, and d_n is presented.     

Gas hold-up and bubble behavior in an upflow packed bed column in the limit of low flow rate

Gas–liquid up-flow in a packed bed is studied experimentally in the limit of zero flow rate. Two patterns of bubble behavior are detected based on the magnitude of the Bond number (Bo). The gas hold-up of the bed is found to be nonmonotonic with a maximum around Bo of unity. The bubble to particle diameter ratio is found to scale inversely with Bo.     

Gas‐Liquid Mass Transfer in Fibrous Bed Reactor with Counter‐Current Liquid Recycle

In this work, the gas‐liquid mass transfer in a lab‐scale fibrous bed reactor with liquid recycle was studied. The volumetric gas‐liquid mass transfer coefficient, k_La, is determined over a range of the superficial liquid velocity (0.0042–0.0126 m.s^–1), gas velocity (0.006–0.021 m.s^–1), surface tension (35–72 mN/m), and viscosity (1–6 mPa.s). Increasing fluid velocities and viscosity, and decreasing interfacial tension, the volumetric oxygen transfer coefficient increased. In contrast to the case of co‐current flow, the effect of gas superficial velocity was found to be more significant than the liquid superficial velocity. This behavior is explained by variation of the coalescing gas fraction and the reduction in bubble size. A correlation for k_La is proposed. The predicted values deviate within ± 15 % from the experimental values, thus, implying that the equation can be used to predict gas‐liquid mass transfer rates in fibrous bed recycle bioreactors.     

Raman spectroscopic study of the microstructure of carbon films developed from cobalt chloride‐modified polyacrylonitrile

Polyacrylonitrile (PAN) was modified with cobalt chloride at 90°C for 5 min. The carbon films prepared from original and modified PAN films were carbonized up to 1300°C. The structure of the resulting carbon film was studied using X‐ray diffraction and Raman spectroscopy. The stacking size obtained from X‐ray diffraction approaches the L_c value of the resulting carbon films as the heat treatment temperature increased. The mean average carbon basal planes in crystalline (Lc/d) also increased with increasing pyrolysis temperature. Raman spectra confirmed the progressive structural ordering as treatment temperature increased. During pyrolysis, a substantial decrease in the intensity of the band near the 1350 cm⁻¹ region was observed, indicating a decrease in the disordered structure. The crystal size (L_a) of the resulting carbon films also showed a remarkable increase with increased heat treatment temperature. The resulting carbon films developed from the modified PAN films had higher L_c and L_a than those developed from the original PAN film. It was established that cobalt catalyzes graphitization of amorphous carbon during pyrolysis. This modification not only promoted the growth of crystal size but also increased the close packing of the carbon basal planes. © 1999 John Wiley & Sons, Inc. J Appl Polym Sci 71: 2219–2225, 1999     

Loading Capacity in Packing Towers – Database,Correlations and Analysis

Experimental results published in the literature between 1935 and 2000 were used to generate a working database of 558 loading capacity data for randomly dumped packed beds. The reported measurements were first used to review the accuracy of the few available predicting loading capacity correlations. The Billet and Schultes semiempirical correlation (Trans IChemE 77 (1999) p. 498) emerged as the best prediction method and is recommended for loading transition estimation, only when the constant C_SO of a given packing element is available. When such a model‐dependent parameter is unavailable, an alternative and generalized neural network correlation is proposed to improve the broadness and accuracy in predicting the loading capacity for packed towers. A combination of five dimensionless groups, namely the liquid Reynolds (Re_L), Galileo (Ga_L) and Stokes (St_L) numbers as well as the packing sphericity (φ) and one bed number (S_B) outlining the tower dimensions were used as inputs of the neural network correlation for the prediction of the loading capacity via the Lockhart‐Martinelli parameter (χ). The correlation yielded an absolute average relative error of 21 % and a standard deviation of 19.9 %. Through a sensitivity analysis, the Stokes number in the liquid phase exhibits the strongest influence on the prediction while the liquid velocity, gas density and packing surface area are the leading physical properties defining the loading level.     

Volumetric Mass Transfer Coefficient in non‐Newtonian Fluids in Spout Fluid Beds

The volumetric liquid‐phase mass transfer coefficient, k_La, was determined by absorption of oxygen in air using six different carboxy‐methyl cellulose (CMC) solutions with different rheological values in three phase spout‐fluid beds operated continuously with respect to both gas and liquid. Three cylindrical columns of 7.4 cm, 11.4 cm, and 14.4 cm diameters were used. Gas velocity was varied between 0.00154–0.00563 m/s, liquid velocity between 0.0116–0.0387 m/s, surface tension between 0.00416–0.0189 N/m, static bed height between 6.0–10.8 cm, and spherical glass particles of 1.75 mm diameter were used as packing material. A single nozzle sparger of 1.0 cm diameter was used in the spouting line. The volumetric mass transfer coefficient was found to increase with gas velocity, liquid velocity, and static bed height and to decrease with the increase of the effective liquid viscosity of the CMC solution. A dimensionless correlation was developed and compared with those listed in the literature.     

Investigation of gas properties, design, and operational parameters on hydrodynamic characteristics, mass transfer, and biomass production from natural gas in an external airlift loop bioreactor

An external airlift loop bioreactor (EALB) was used for production of biomass from natural gas. The effect of riser to downcomer cross sectional area ratio (A_r/A_d), volume of gas-liquid separator, superficial gas velocity (U_sgr), and physical properties of gases and their mixtures [υ_g (μ/ρ) and D_g] were investigated on mixing time, gas hold-up, and volumetric gas liquid mass transfer coefficients (k_La). It was found that A_r/A_d has remarkable effects on gas hold-up and k_La due to its influence on mixing time. Kinematic viscosity (υ_g) showed its significant role on mixing time, gas hold-up and k_La when different gases used (mixing time changes directly whereas gas hold-up and k_La change indirectly). Moreover, it was found that diffusion coefficient of gas in water (D_g) has remarkable effect on k_La. The volumetric mass transfer coefficients for methane and its mixtures with oxygen (three different mixtures) were determined at different geometrical and operational factors. In average, the rate of oxygen utilization is approximately 1.8 times higher than that of methane. A gas mixture of 25 vol% methane and 75 vol% oxygen was the best gas mixture for biomass production in the EALB. The correlations developed for predicting the mixing time, gas hold-up, and k_La in terms of U_sgr, A_r/A_d, volume of gas-liquid separator, and gas phase properties have been found to be encouraging.     

A three‐phase fluidized bed reactor in the combined anaerobic/aerobic treatment of wastewater

Aerobic degradation or polishing is an essential step in the combined anaerobic/aerobic treatment of wastewater. In this study, a type of porous glass beads was used for immobilization of microbial cells in a three‐phase aerobic fluidized bed reactor (AFBR) with an external liquid circulation. The effects of superficial gas and liquid velocities on bed expansion, solid and gas hold‐ups and specific oxygen mass transfer rate, k_La, were investigated. A tracer study showed that the mixing and flow pattern in the 8 dm³ reactor could be simulated by a non‐ideal model of two continuous stirred tank reactors (CSTRs) in series. By treating an effluent from an upflow anaerobic sludge blanket (UASB) digester, the distribution of suspended and immobilized biomass in the reactor as well as the kinetics of COD removal were determined. The specific oxygen mass transfer rate, k_La, at a superficial gas velocity of 0.7 cm s⁻¹ dropped by about 30% from 32 h⁻¹ in tap water to 22 h⁻¹ after a carrier load of 15% (v/v) was added. The measured k_La further dropped by about 20% to 18 h⁻¹ in the wastewater, a typical value of the bubbling fermenters with no stirring. Compared with the aerobic heterotrophs under optimum growth conditions, the microbes in this reactor which was fed with anaerobic effluent plus biomass behaved like oligotrophs and showed slow specific COD removal rates. This might be attributed to the presence of a significant amount of obligate anaerobes and facultative organisms in the aerobic reactor. This was confirmed by a relatively low intrinsic oxygen uptake rate of the microbial population in the reactor, 94 mg O₂ dm⁻³ h⁻¹ or 19 mg O₂g VS⁻¹ h⁻¹. © 1999 Society of Chemical Industry     

Interfacial Area and Liquid‐Side Volumetric Mass Transfer Coefficient in a Downflow Bubble Column

An experimental investigation was made to measure interfacial area, a, and liquid‐side volumetric mass transfer coefficient, k_La, in a downflow bubble column by chemical methods viz., absorbing CO₂ in aqueous sodium hydroxide and sodium carbonate/bicarbonate buffer solution respectively. The effect of gas and liquid flowrate and nozzle sizes on a and k_La were investigated. The experimental data obtained in the present system were analyzed and correlations were developed to predict a and k_La in terms of superficial gas velocity. The variation of a and k_La with specific power input were shown in graphical plot and compared with other gas‐liquid systems.     

Experimental study of hydromechanics and mass transfer characteristics in cross flow packed tower

The functions of hydromechanics and volumetric mass transfer coefficient K_Xa of cross flow packed tower were studied by setting baffle plates. The influence of plate spacing on the pressure drop, height of the mass transfer unit, and mass transfer characteristics were investigated in an ordinary packed tower and a cross flow packed tower. The pressure drop increased with a rise in the flow rate of gas and liquid, and an excessive pressure drop caused flooding in the cross flow packed tower. The height of the mass transfer unit increased with a decrease in the gas flow when H/D = 0.8 ? 1.2, while increased with the increase of the gas flow when H/D = 0.6. In the ordinary packed tower and the cross flow packed tower, K_Xa increased with a rise of liquid flow. The influence of gas flow on K_Xa was negligible in the ordinary packed tower, however, in the cross flow packed tower the K_Xa gradually increased with a rise in the gas flow rate. The effect of mass transfer was optimal at H/D = 0.8. In addition, using STATISTICA software, the corresponding K_Xa correlations were proposed and discussed.     

Space‐charge limited conduction in polyaniline films

Polyaniline emeraldine base (PANI‐EB) powder was synthesized by oxidative polymerization of aniline. The PANI‐EB films were prepared by the solution‐casting technique. The temperature‐dependent dc conductivity measured in the range 173–303 K suggests that the PANI‐EB is a quasi‐one‐dimensional disordered conductor. The current‐voltage characteristics of the PANI‐EB films measured in the range 333–383 K showed the SCLC mechanism. The SCLC parameters such as free carrier density (p₀), trap density (p_t), the ratio between free carrier density to the total carrier density (θ), mobility (µ) and the effective hole mobility (µ_eff) were calculated. The activation energy (E_a = 0.32 eV) and the Fermi level (E_F = 0.42 eV) were estimated. As well as these, the trap parameters such as the trap filled limit voltage (V_TFL), the shallow trap density (N_t), the depth of the dominant trap level (E_t ? E_v), the density of states within the hole mobility edge (N_v) and the characteristic energy (E_c) were also calculated and presented. The exponential type of traps distribution with large number of traps was found to be due to the disorder and moisture in the polymer films. Copyright © 2004 Society of Chemical Industry     

EFFECT OF STATIC LIQUID HEIGHT ON GAS-LIQUID MASS TRANSFER IN A DRAFT-TUBE BUBBLE COLUMN AND THREE-PHASE FLUIDIZED BED

Experiments are performed under batch-liquid operating conditions to investigate the effect of static liquid height on the gas-liquid mass transfer coefficient (K_La) in a draft-tube bubble column (DTBC) and a draft-tube three-phase fluidized bed (DTFB). In addition, the effects of column diameter, gas-distributor, and draft-tube diameter are studied. The results indicate that for a given system with a porous plate gas-distributor at low superficial gas velocities (<70 m/hr), increasing static liquid height decreases K_La. At high gas velocities, K_La is independent of the static liquid height. For systems with a perforated gas-distributor, there is no effect of static liquid height on K_La. The formation of small dispersed bubbles at low gas velocities in the porous plate distributor system accounts for the considerably high K_La values and the observed effect of liquid height. On the other hand, the formation of large spherical-cap bubbles and the bubble coalescence at high gas velocities reduce the performance of the porous plate distributor system to that of the perforated one.     

A comparison of triglyceride oil hydrogenation in a downflow bubble column using slurry and fixed bed catalysts

The hydrogenation of the triglyceride oil, soya bean oil, has been studied in the temperature range 130–160 °C and in the pressure range 100–600 kPa using (i) a 5% w/w Pd/C slurry catalyst and (ii) a 3% w/w Pd/Al₂O₃ Raschig ring catalyst in a cocurrent downflow contactor (CDC) reactor. Separate studies of residence time distribution (RTD) were carried out in a modified CDC device in order to determine dispersion numbers and dispersion coefficients. The RTD measurements indicated that the overall flow was a mixture of well‐mixed and plug flow for the unpacked CDC, so that the entry section (0–30 cm from entrance) was perfectly mixed and the remainder of the column (30–130 cm) gave predominantly plug flow behaviour. The introduction of random packing in the form of 13 mm Raschig rings gave rise to increased back mixing in the lower part of the CDC and the overall dispersion number increased due to liquid and gas circulation around the packing elements. Kinetic studies revealed an initial rate reaction order of 1.24–1.26 with respect to hydrogen concentration both in slurry and fixed bed CDC reactors and is interpreted as a combination of a parallel pair of first and second order reactions during the initial stages of reaction. Mass transfer coefficients for gas absorption (k_La) and liquid–solid mass transport (k_s) were determined for both types of reactor. The k_La values lay in the range 1.0–3.33 s⁻¹ and the liquid–solid transport resistances (X_LS) were all <1%, so that the reaction was almost totally surface reaction rate controlled. Apparent energy of activation measurements gave values of E_A = 49 ± 6 kJ mol⁻¹, which is strongly indicative of surface reaction rate control involving the hydrogenation of an olefinic double bond. The selectivity in respect of linolenate (three double bonds) removal and linoleate (two double bonds) retention was high with, for palladium, relatively low trans‐isomer production (<30%). The overall selectivity was slightly, but significantly, better for the fixed bed CDC reactor and this is attributed to the greater degree of plug flow behaviour in the latter, despite the bed causing an increase in dispersion number. However, there is no reaction in the well‐mixed section of the fixed bed CDC reactor as there is in the slurry CDC reactor and this is likely to improve selectivity in a consecutive reaction sequence. © 2000 Society of Chemical Industry